脑胶质瘤糖原代谢对其恶性生物学行为的影响研究  

The impact of glycogen metabolism on the malignant biological behaviors of brain gliomas

作　　者：李文昊 贺御泽 舒雨婷 杨万纯 刘艳辉[1] Li Wenhao;He Yuze;Shu Yuting;Yang Wanchun;Liu Yanhui(Department of Neurosurgery,West China Hospital,Sichuan University,Chengdu 610041,China;State Key Laboratory of Biotherapy,Sichuan University,Chengdu 610041,China)

机构地区：[1]四川大学华西医院神经外科,成都610041 [2]四川大学生物治疗全国重点实验室,成都610041

出　　处：《中华神经外科杂志》2025年第3期295-301,共7页Chinese Journal of Neurosurgery

基　　金：国家自然科学基金(82272644);四川省科技计划(2023YFQ0002)。

摘　　要：目的初步探讨脑胶质瘤糖原代谢对其恶性生物学行为的影响。方法对四川大学华西医院神经外科手术切除的2例脑胶质瘤(高级别和低级别胶质瘤各1例)患者的胶质瘤样本进行过碘酸-雪夫染色,明确不同级别胶质瘤组织中的糖原分布。对四川大学华西医院收治的脑胶质瘤患者(简称华西队列,共83例)、癌症基因组图谱(TCGA)数据库中的多形性胶质母细胞瘤及低级别胶质瘤患者(简称TCGA队列,共624例)和中国脑胶质瘤基因组图谱(CGGA)数据库中的mRNAseq_325队列(简称CGGA队列,共221例)进行数据提取和分析。将TCGA队列按6∶4的比例分为训练集(375例)和内部验证集(249例),CGGA和华西队列作为外部验证集。在TCGA队列训练集中,应用最小绝对收缩和选择运算符-Cox回归模型从糖原代谢相关基因(GMGs)中筛选出糖原代谢的关键基因,构建GMGs风险评分(GMRS)模型,即关键基因的相对表达量与其回归系数乘积之和。在内部和外部验证集中,使用R软件"survminer"包根据GMRS模型的最优截断值将患者分为GMRS模型高风险组和低风险组,在每个验证集中比较两组患者的生存预后。通过单因素和多因素Cox回归模型,判断GMRS模型高风险组是否为脑胶质瘤患者生存预后的独立危险因素。使用受试者工作特征(ROC)曲线评价GMRS模型对胶质瘤患者6、12、18个月生存预后的预测效能;ROC曲线的曲线下面积(AUC)>0.5为模型具有预测作用。取人胶质母细胞瘤细胞株U87,加入糖原磷酸化酶抑制剂(10μmol/L)共培养的为实验组,未加入的为对照组(各组n=3);采用细胞计数试剂盒(CCK-8)增殖和Transwell迁移实验检测两组细胞的增殖和迁移能力。结果过碘酸-雪夫染色结果显示糖原显著富集于胶质瘤细胞胞体中。在TCGA队列训练集中,筛选出关键基因HK3、AKT2、PGM1、UBB、PPP2R5D、HKDC1、EPM2A和PHKG2用于构建GMRS模型;多因素Cox回归模型分析显示,ObjectiveTo preliminarily investigate the impact of glycogen metabolism in brain gliomas on their malignant biological behaviors.MethodsGlioma samples from two patients(one with high-grade and one with low-grade glioma)who underwent surgical treatment at the Department of Neurosurgery,West China Hospital,Sichuan University were subjected to periodic acid-Schiff(PAS)staining to identify glycogen distribution in glioma tissues of different grades.Data extraction and analysis were performed on brain glioma patients from West China Hospital(West cohort,83 cases),the Cancer Genome Atlas(TCGA)dataset including glioblastoma and low-grade gliomas(TCGA cohort,624 cases),and the Chinese Glioma Genome Atlas(CGGA)mRNAseq_325 dataset(CGGA cohort,221 cases).The TCGA cohort was divided into a training set(375 cases)and an internal validation set(249 cases)in a 6∶4 ratio,while the CGGA and West cohorts served as external validation sets.In the TCGA training set,key genes related to glycogen metabolism were identified using the Least Absolute Shrinkage and Selection Operator-Cox regression model.Glycogen metabolism related genes risk score(GMRS)was the sum of the products of the expression levels of key genes and their corresponding regression coefficients.In both internal and external validation sets,patients were categorized into high-risk and low-risk groups based on the optimal cutoff value of the GMRS model using the"survminer"package in the R software,and survival outcomes between these two groups were compared.Univariate and multivariate Cox regression models were used to assess whether the high-risk GMRS model group was an independent risk factor for survival prognosis in glioma patients.The predictive performance of the GMRS model for 6-,12-,and 18-month survival prognosis was evaluated using receiver operating characteristic(ROC)curves,with an area under the curve(AUC)>0.5 indicating predictive validity.In vitro experiments were performed using the human glioblastoma cell line U87,with the experimental group treated w

关 键 词：神经胶质瘤 糖原 代谢 基因表达调控 细胞迁移分析 细胞增殖 模型 统计学 

分 类 号：R739.41[医药卫生—肿瘤]